Advances in Automated Neonatal Seizure Detection

This chapter highlights the current approaches in automated neonatal seizure detection and in particular focuses on classifier based methods. Automated detection of neonatal seizures has the potential to greatly improve the outcome of patients in the neonatal intensive care unit. The electroencephalogram (EEG) is the only signal on which 100% of electrographic seizures are visible and thus is considered the gold standard for neonatal seizure detection. Although a number of methods and algorithms have been proposed previously to automatically detect neonatal seizures, to date their transition to clinical use has been limited due to poor performances mainly attributed to large inter and intra-patient variability of seizure patterns and the presence of artifacts. Here, a novel detector is proposed based on time-domain, frequency-domain and information theory analysis of the signal combined with pattern recognition using machine learning principles. The proposed methodology is based on a classifier with a large and diverse feature set and includes a post-processing stage to incorporate contextual information of the signal. It is shown that this methodology achieves high classification accuracy for both classifiers and allows for the use of soft decisions, such as the probability of seizure over time, to be displayed.

[1]  Robert R. Clancy,et al.  Characterization of neonatal seizures by conventional EEG and single-channel EEG , 2007, Clinical Neurophysiology.

[2]  Brian Litt,et al.  One-Class Novelty Detection for Seizure Analysis from Intracranial EEG , 2006, J. Mach. Learn. Res..

[3]  J. Tyson,et al.  Risk factors for neonatal seizures: a population-based study, Harris County, Texas, 1992-1994. , 2001, American journal of epidemiology.

[4]  B. Hjorth.  EEG analysis based on time domain properties. , 1970, Electroencephalography and clinical neurophysiology.

[5]  T. Inder,et al.  Seizure detection algorithm for neonates based on wave-sequence analysis , 2006, Clinical Neurophysiology.

[6]  David G. Stork,et al.  Pattern Classification , 1973 .

[7]  R. Grebe,et al.  Automated neonatal seizure detection: A multistage classification system through feature selection based on relevance and redundancy analysis , 2006, Clinical Neurophysiology.

[8]  Nurettin Acir,et al.  Automatic spike detection in EEG by a two-stage procedure based on support vector machines , 2004, Comput. Biol. Medicine.

[9]  Saman K. Halgamuge,et al.  Computational Intelligence for Modelling and Prediction , 2005, Studies in Computational Intelligence.

[10]  José del R. Millán,et al.  Person Authentication Using Brainwaves (EEG) and Maximum A Posteriori Model Adaptation , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Yuko Mizuno-Matsumoto,et al.  Wavelet analysis for neonatal electroencephalographic seizures. , 2003, Pediatric neurology.

[12]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[13]  G B Boylan,et al.  Non-expert use of the cerebral function monitor for neonatal seizure detection , 2004, Archives of disease in childhood. Fetal and neonatal edition.

[14]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[15]  V. Vapnik.  Estimation of Dependences Based on Empirical Data , 2006 .

[16]  A. W. de Weerd.  Atlas of EEG in the first months of life , 1995 .

[17]  Shigeo Abe DrEng.  Pattern Classification , 2001, Springer London.

[18]  L. Duchene,et al.  An Optimal Transformation for Discriminant and Principal Component Analysis , 1988, IEEE Trans. Pattern Anal. Mach. Intell..

[19]  Douglas A. Reynolds,et al.  Speaker Verification Using Adapted Gaussian Mixture Models , 2000, Digit. Signal Process..

[20]  Brian Litt,et al.  Epileptic seizure prediction using hybrid feature selection over multiple intracranial EEG electrode contacts: a report of four patients , 2003, IEEE Transactions on Biomedical Engineering.

[21]  S. Huffel,et al.  Automated neonatal seizure detection mimicking a human observer reading EEG , 2008, Clinical Neurophysiology.

[22]  L. Tippett,et al.  Applied Statistics. A Journal of the Royal Statistical Society , 1952 .

[23]  G. Lightbody,et al.  A comparison of quantitative EEG features for neonatal seizure detection , 2008, Clinical Neurophysiology.

[24]  R B Reilly,et al.  Classifier models and architectures for EEG-based neonatal seizure detection , 2008, Physiological measurement.

[25]  Ali H. Shoeb,et al.  Impact of Patient-Specificity on Seizure Onset Detection Performance , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[26]  J. Gotman,et al.  Automatic seizure detection in the newborn: methods and initial evaluation. , 1997, Electroencephalography and clinical neurophysiology.

[27]  A M Bye,et al.  Spatial and Temporal Characteristics of Neonatal Seizures , 1995, Epilepsia.

[28]  Truong Q. Nguyen,et al.  Gaussian mixture models of ECoG signal features for improved detection of epileptic seizures. , 2004, Medical engineering & physics.

[29]  Scott B. Wilson,et al.  Seizure detection: correlation of human experts , 2003, Clinical Neurophysiology.

[30]  A. Aarabi,et al.  A multistage knowledge-based system for EEG seizure detection in newborn infants , 2007, Clinical Neurophysiology.

[31]  Alexander J. Smola,et al.  Advances in Large Margin Classifiers , 2000 .

[32]  J. Frost,et al.  A Multistage System for the Automated Detection of Epileptic Seizures in Neonatal Electroencephalography , 2009, Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society.

[33]  Ali H. Shoeb,et al.  Patient-specific seizure onset detection , 2004, Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[34]  Alexander J. Smola,et al.  Learning with kernels , 1998 .